NK cells and T cell express innate NK receptors such as NKG2D, which enable the cells to recognize and eliminate transformed cells. The NKG2D receptor, which activates target cell cytolysis and cytokine production by NK cells, binds to 5 or more independent cell surface ligands (NKG2D ligands). NKG2D ligands are upregulated in tumor cells. Once a cell expresses cell surface NKG2D ligands, NK cells can kill it. In cancer patients and some other syndromes, it has been shown that some human NKG2D ligands are cleaved from the cell surface by proteases and accumulate in serum. It is believed that the shed ligands can interfere with NKG2D recognition and prevent tumor cell elimination, but we believe that much of the evidence for this proposition is flawed and incomplete. Direct testing of it has been hampered by the lack of an animal model. We have filled this gap by showing that a mouse NKG2D ligand, MULT1, is cleaved efficiently from cells by proteases, and accumulates at high concentrations in the blood of mice with tumors and inflammatory diseases. Cleaved MULT1 binds to NKG2D with high affinity (~10 nM). To address the biological effects of a shed form of MULT1, tumor cells that did not express endogenous NKG2D ligands were engineered to produce a secreted form of MULT1 similar to the shed form. Whereas we expected that secreted MULT1 might interfere with tumor rejection responses, we found instead that tumor cells secreting MULT1 were rejected. Tumor rejection was mediated by NK cells, and in some cases T cells, and required that the host express NKG2D. These findings prompted experiments that showed that provision of purified recombinant MULT1 along with B16 melanoma cells resulted in induction of NK function ex vivo and tumor rejection in vivo. The data as a whole support a model in which expression of NKG2D ligands by non-tumor cells, such as tumor associated macrophages, persistently stimulate NK cells, resulting in desensitization of the NK cells; interruption of the interaction by soluble MULT1 prevents these desensitizing interactions, and therefore restores NK cell functionality and tumor rejection via recognition of distinct tumor cell ligands for NK cells. Remarkably, preliminary data show that injection of a blocking NKG2D antibody enhanced the responsiveness of NK cells tested a few days later, ex vivo, supporting the model. Thus, whereas it is widely assumed that shed NKG2D ligands inhibit anti-tumor immune responses, our data indicate that shed MULT1 mobilizes NK cell responses against tumors. These findings have important mechanistic implications and at the same time suggest that soluble NKG2D ligands, or antibodies that block the interaction, have considerable potential as immunotherapeutic agents. We propose the following specific aims: (Aim 1) To document immune-enhancement by soluble ligands and antibodies to NKG2D or its ligands; (Aim 2) To determine the mechanisms of immune-enhancement and tumor rejection; (Aim 3) To assess the immunotherapeutic potential of soluble ligands and antibodies.